Journal of Comparative Physiology A (2020) 206:543–552 https://doi.org/10.1007/s00359-020-01424-8 ORIGINAL PAPER Hearing and sound localization in Cottontail rabbits, Sylvilagus foridanus Rickye S. Hefner1 · Gimseong Koay1 · Henry E. Hefner1 Received: 4 April 2020 / Revised: 5 May 2020 / Accepted: 14 May 2020 / Published online: 1 June 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020 Abstract Cottontail rabbits represent the frst wild species of the order of lagomorphs whose hearing abilities have been determined. Cottontails, Sylvilagus foridanus, evolved in the New World, but have spread worldwide. Their hearing was tested behavio- rally using a conditioned-avoidance procedure. At a level of 60 dB SPL, their hearing ranged from 300 Hz to 32 kHz, a span of 7.5 octaves. Mammalian low-frequency hearing is bimodally distributed and Cottontail rabbits fall into the group that hears below 400 Hz. However, their 300-Hz limit puts them near the gap that separates the two populations. The minimum audible angle of cottontails is 27.6°, making them less acute than most other species of mammals. Their large sound-localization threshold is consistent with the observation that mammals with broad felds of best vision require less acuity to direct their eyes to the sources of sound. Keywords Behavioral audiogram · Low-frequency hearing · Comparative hearing · Sound localization and vision · Animal psychophysics Introduction 2007). Both the audiogram and noise-localization thresholds of three Eastern cottontails were determined for comparison The Order Lagomorpha, with approximately 90 species, with those of other mammals. includes rabbits, hares, and pikas (Melo-Ferreira and Alves 2018). Currently, the only available audiogram for this Orycto- Order is that of the domesticated Old-World rabbit, Methods lagus cuniculus, a burrowing species (Hefner and Master- ton 1980). To extend our survey of mammalian hearing to include a non-domesticated species in this group, we report The rabbits were tested using a conditioned-avoidance pro- here the audiogram of the Eastern cottontail rabbit (Sylvila- cedure in which a thirsty animal was trained to maintain gus foridanus). mouth contact with a water spout to receive a steady trickle The Cottontail rabbit is a member of a New-world genus of water. Warning sounds were presented intermittently, fol- native to much of North and Central America, including the lowed at their ofset by a mild electric shock delivered via northern parts of South America (Chapman et al. 1980). the spout. The animals learned to avoid the shock by break- It is a smaller non-burrowing species, nesting in slight ing contact with the spout when they heard a warning sound. depressions, that can be compared to the larger domesti- The audiogram was determined for pure tones ranging cated representatives of Old-world burrowing rabbits (e.g., from 125 Hz to 64 kHz. Sound-localization acuity (mini- New Zealand White and Dutch Belted) that have served as mum audible angle) was determined for 100-ms broadband models for mammalian sound localization (e.g., Blanks et al. noise pulses centered left and right on the midline in the azimuthal plane. * Rickye S. Hefner [email protected] 1 Department of Psychology, University of Toledo, Toledo, OH 43606, USA Vol.:(0123456789)1 3 544 Journal of Comparative Physiology A (2020) 206:543–552 Subjects Finally, a 15-W light was mounted approximately 0.5 m below the cage and was turned on and of simultaneously Three Eastern cottontail rabbits, S. foridanus (two females with the shock to indicate to the animal when a shock had A, B, and one male C), were wild trapped in Lucas County, been delivered, and when it was safe to return to the spout Ohio, and maintained in the laboratory. They were approx- at the end of successful detection trials. imately 6 months old and weighed 918–1099 g on ad libi- tum feed at the beginning of testing. They were housed in stainless steel cages (61 × 46 × 70 cm) and given free Acoustical apparatus access to rabbit chow supplemented by occasional fruits and vegetables. Audiogram While on test, the animals received their water only in the test sessions and were weighed daily to monitor their Pure tones from 125 Hz to 64 kHz were produced using a deprivational state. During testing, they maintained at signal generator (Krohn-Hite 2400 AM/FM) and were con- least 80% of their ad libitum weights and returned to those tinuously verifed by a frequency counter (Fluke 1900A). weights within a few days of ad libitum food and water. The signal was shaped by a rise/decay gate (Coulbourn S84-04) allowing 10 ms rise/decay times for all frequencies of 1 kHz and higher. Longer rise/decay times were used Behavioral apparatus at lower frequencies to allow the signal to reach full volt- age (and fall to zero voltage) over at least ten cycles. For All testing was carried out in a double-walled chamber the audiogram, pure tones were presented as four pulses of (IAC model 1204, 2.55 × 2.75 × 2.05 m). To reduce sound 400-ms duration with 100 ms between pulses. The intensity refection, the foor was carpeted and the walls and ceiling of the tones was adjusted in 5-dB steps using an attenuator were lined with egg crate foam. All acoustic and behav- (Hewlett Packard 350D), the linearity of which was cali- ioral equipment was located adjacent to the chamber and brated throughout the voltage range used for the diferent the rabbits were observed over closed-circuit television. intensities being tested. The electrical signal was then band- The test cage measured 55 × 31 × 38 cm and was con- pass fltered (Krohn-Hite 3550; ± 1/3 octave) to reduce any structed of 1-in (2.54-cm) welded wire mesh. The legs possible electrical noise and routed to an amplifer (Crown supporting the test cage were placed on 8-cm-thick foam D75). Output from the amplifer to speaker was monitored pads as a precaution against substrate-borne vibrations. In for distortion and noise with an oscilloscope. Loudspeakers the front of the cage, a water spout protruded through the were placed at ear level 1 m in front of the animal’s ears. A foor to a comfortable drinking height. The spout consisted variety of loudspeakers, depending on the frequencies being 3 of 15-gauge stainless steel tubing with a 1.5 × 2.5 cm stain- tested, was used: 15-in (38-cm) woofer in a 0.45 m enclo- less steel oval welded to the tip serving as a lick surface. sure, piezoelectric tweeters (Motorola KSN 1005A), and rib- The tip of the spout protruded 5 cm above the cage foor, bon tweeters (Foster E110T02). The woofer used to generate below the level of the animals’ ears, thus minimizing the lowest frequencies was placed on 8-cm-thick foam pads obstructions between the ears and the loudspeakers. The as a precaution against substrate-borne vibrations. Some fre- water spout was connected via plastic tubing to a 50 mL quencies were tested using more than one of the speakers. 2 syringe pump (Thompson et al. 1990) located outside the The sound pressure level (SPL re 20 µN/m ) was meas- test chamber. The pump supplied a slow trickle of water ured daily using a Brüel and Kjaer (B&K) 1/4-in (0.64-cm) as long as an animal maintained contact with the spout. microphone, coupled with a preamplifer (B&K 2618), a The water delivery rate was adjusted, so that the animals microphone amplifer (B&K 2608), and spectrum analyzer could obtain their daily water in a single test session last- (Zonic 3525). The measuring system was calibrated with a ing 35–65 min. Requiring the animals to keep their mouths pistonphone (B&K 4230). Sound measurements were taken on the water spout served to keep their heads in a fxed by placing the microphone in the position occupied by an position relative to the loudspeakers. A contact circuit, animal’s head when it was drinking and pointing the micro- connected between the spout and cage foor, detected when phone directly toward the loudspeaker (0° incidence). Care a rabbit made contact with the spout and activated the was taken to produce a homogeneous sound feld (± 1 dB) in syringe pump. In addition, a shock generator was con- the area occupied by the animal’s head and ears. The tones nected between the spout and the cage foor to provide produced by the loudspeakers during testing were examined feedback and a mild cost for failing to respond to warning for the presence of overtones or distortion by routing the sounds. The shock (0.3 s) was adjusted for each rabbit to output of the microphone amplifer directly to the spectrum the lowest level that elicited a reliable avoidance response. analyzer (Zonic 3525). Background noise levels and spectra were also obtained in this manner. Analysis of the fnal test 1 3 Journal of Comparative Physiology A (2020) 206:543–552 545 signals indicated that any overtones were more than 10 dB Testing continued until a rabbit no longer responded below the animals’ thresholds. to the warning signal above the level expected by chance, i.e., the hit rate was no longer signifcantly higher than the Sound localization false-alarm rate (p > 0.05, binomial distribution). A typical test session for a trained animal consisted of approximately To determine minimum audible angle, a single 100-ms 35–75 warning trials and approximately four times as many broadband noise burst (2–45 kHz) was emitted from one of safe (silent) trials during which a complete psychophysical ten piezoelectric tweeters (fve pairs with closely matched curve could be generated.